back

»With the KI-Navi Handel, we show how you can easily use artificial intelligence yourself without the need for a large IT-department. We explain how existing solutions that are relevant to the retail sector can be implemented.«

Frank Rehme is an innovator and expert on the challenges facing the retail industry. His career has been anything but linear; he literally started at the bottom: 45 years ago, Rehme started out in mining. After working in engineering and IT, he finally came to retail. He founded the gmvteam GmbH in 2013 and supports retail and industry initiatives with his company. In 2024, he and his company took over the consortium leadership of the KI.NRW flagship project KI-Navi Handel. We asked him what is behind the project, where he sees opportunities with AI for the retail sector and what the goals of the project are.

Mr. Rehme, you have many years of professional experience in the retail industry, what is your assessment: how open is the industry to technological progress through AI?
The retail sector was one of the pioneers in the field of digitalization. Many processes were already IT-supported in the 1970s. The first sentence I heard in the industry was: “IT and logistics are the backbone of retail”. The topics of automation and e-commerce shaped the years that followed. Now the industry is facing new challenges: a shortage of skilled workers, disrupted supply chains and changing customer needs. All of these challenges require further steps that can be solved by AI. It is therefore not necessary to create an openness in the industry – it is already there!

The AI hype created by language models such as ChatGPT has spread to many sectors. Nevertheless, you say in the KI-Navi Handel project that the use of AI in retail has so far been limited. Why is that?
The assessment primarily relates to SMEs, which have so far had little opportunity to engage with the topic. We’re talking about more than 250,000 stores across Germany, mostly owner-managed, with a turnover of less than two million euros. These companies need support, and that’s what the KI-Navi Handel project is for. We all benefit from this support, because it is precisely these stores that make our city centers unique, beyond the large chain stores.

How do you want to change that? What solutions are you working on to make AI more popular in retail? What does the name KI-Navi mean?
We deliberately chose the navigation system analogy. When I’m traveling in an unfamiliar area, the navigation system guides me to my destination. In this project, we show how you can easily use artificial intelligence yourself without the need for a large IT-department. We explain how existing solutions relevant to retail can be implemented. We have shown this in the area of chatbots, automated social media work and analyzing customer movements. Almost all solutions that can be implemented in 30 minutes.

Where do you see further AI potential for the retail sector?
Particularly in the area of automation in the administrative sector. Reporting, HR management and digital document management are particularly in demand. In times of a shortage of skilled workers, staff should be relieved of these tasks so that more focus can be placed on the needs of customers. Another important area is the optimization of offer communication. Knowing at an early stage what people want and aligning the supply chain and marketing accordingly is a success factor for the future.

Which data sources do you use to develop AI solutions? And how good is data availability in the industry?
This is a hot topic in all AI projects. We are talking about a magic triangle, as I like to call it: on the one hand, data availability, then data quality and, thirdly, we mustn’t forget data protection. The latter is an important issue, but is sometimes exaggerated and puts off small retailers in particular. In general, we can say that well-maintained data from merchandise management systems and loyalty programs form the first basis.

How do small retailers differ from large retail chains?
Small companies excel first and foremost in the area of specialization. They know their customers very well and have great flexibility to react to changing conditions. However, the large companies have the advantage of being able to base their pricing on different conditions due to their purchasing power and large volumes. This combination can be seen in vibrant city centers: the big ones provide the magnet effect, the small ones the individuality.

Which use cases do you support?
In the evaluation of sales histories, which is called associative shopping basket analysis. We evaluate which follow-up purchases are made with what frequency, e.g. after a barbecue has been purchased. Exactly these items are then communicated to the right target group in customized and AI-supported communication measures. Another topic is customer relationship management: how can I interact better with customers and communicate with a high level of relevance? A new generation of chatbots that work much more flexibly and precisely than before is helping us here.

The project of the state of NRW is initially set to run for four years. Where do you hope the retail sector will be in four years’ time with regard to the use of AI?
If the majority of the mentioned target group understands the acronym AI not as “not interested” but as “can do”, then we are on the right track. I am confident of this, because current developments are catapulting this technology into the center of society.